← Back to People
Surya Dantuluri authored an article on how every serious AI company will eventually train its own models as barriers to model training collapse.
3
Mentions
3
Articles
3
Outlets
Aug 2025 — Feb 2026
Coverage
Associated Topics
Companies Covered
Coverage Patterns
How media typically covers Surya Dantuluri
Article Types
- announcement1
- tutorial1
Media Angles
- product1
- technical1
Narrative Framing
- breakthrough2
Appears In
Associated AI Models
GPT-51GPT-5-mini1prava-fc-small1Kimi K21
Writes About
Articles
Most recent firstArticles Written
Surya Dantuluri as author
Teaching LLMs to Be Funny
jokegen.sdan.iotutorialpositiveFeb 3, 2026
Using rubric-based reinforcement learning to decompose subjective qualities like humor into verifiable properties enables LLMs to generate funnier, more contextually relevant jokes.
“Author of "Teaching LLMs to Be Funny"”
LLMsAI ResearchReinforcement LearningGenerative AI
Also Mentioned In
Referenced in coverage
The Training Imperative
sdan.substack.comcautiousSep 10, 2025
“Surya Dantuluri authored an article on how every serious AI company will eventually train its own models as barriers to model training collapse.”
Teaching GPT-5 to Use a Computer
prava.coannouncementpositiveAug 18, 2025
GPT-5's advanced reasoning capabilities enable building a computer copilot (Archon) that can understand natural language instructions and autonomously control computer interfaces through vision and action planning.
“Author and hackathon winner who built Archon, a GPT-5 powered computer control system”
AI AgentsFoundation ModelsGenerative AIAI Products